So I've been playing with a design using old soviet germanium transistors and running into a stability problem. This is the schematic that I prototyped with all values the same except using a P602 transistor for the VAS stage. The instability happens when the voltage is increased to a certain level, say 20 volts. It breaks into oscillation and the current spikes along with lower output transistor getting warm. With a speaker it sounds like static and looks like random noise on the scope. I tried 100 ohm base resistors on the output devices and large and small power decoupling caps but nothing makes a difference. Finally managed to catch the start of the oscillation with my scope across the base resistor of the lower output transistor. Who knew these slow leaky GE devices would be oscillating at a few Mhz.
Attachments
I would opt for an avalanche-like breakdown. To be sure, replace the possible culprit with a modern Si one (correct the bias if required).
If it holds for much more than 20V, you know where your problem is.
I have tried to make amplifiers based on Ge devices, but the breakdown voltage was a big problem
If it holds for much more than 20V, you know where your problem is.
I have tried to make amplifiers based on Ge devices, but the breakdown voltage was a big problem
That's a good theory and explains why nothing is really visible on the scope after the breakdown. The date codes on the output transistors range from 1984 to 1990 and if you buy 10 you'll likely get 2-3 "good" ones from the batch. A buddy of mine cut open a bad one and the construction looks decent. The can is filled with some kind of clear silicone grease and the die has a blob of what feels like clear RTV over it. So my guess manufacturing impurities. The die in the pic is scratched up from scraping away the coating.
I have some AD140 transistors that don't exhibit this behavior but even after playing with the bias their THD was much higher than the soviets. Like 2-3% compared to 0.3%. Which seems odd as the Leak Stereo 30 used them.
Attachments
GT806 (AL102, AUY35)
fгр, f*h21б, f**h21э, f***max
≥10*MHz
https://rudatasheet.ru/transistors/gt806/
AD140
0.2 MHz
https://alltransistors.com/ru/transistor.php?transistor=21541
When used in class A, germanium transistors have no advantage over silicon ones (Cooling problems). Mosfet works well.
The input transistor operates in microcurrent mode, so it is better to use silicon.
fгр, f*h21б, f**h21э, f***max
≥10*MHz
https://rudatasheet.ru/transistors/gt806/
AD140
0.2 MHz
https://alltransistors.com/ru/transistor.php?transistor=21541
When used in class A, germanium transistors have no advantage over silicon ones (Cooling problems). Mosfet works well.
The input transistor operates in microcurrent mode, so it is better to use silicon.
Last edited:
Yeah I know silicon is superior but we do have the vacuum tube forum after all 😛
Elvee was correct about the avalanche breakdown. Tacked in a pair of TIP2955 transistors and no problems all the way up to 50 volts.
Elvee was correct about the avalanche breakdown. Tacked in a pair of TIP2955 transistors and no problems all the way up to 50 volts.